Exam 3 Study Guide

These three chapters tell one story from three angles. A gene changes, a drug stops working, and the only answers left are good technique and good drug choices. Read the "what to know" prose first. Then cover the right side of each table and rebuild it. Say the sequences out loud in order, because this exam rewards knowing what comes next. Finish each chapter with its ten self-test items. The cram sheet at the end gathers every drug, number and rule in one place.

Chapter 7Microbial Genetics and How Resistance Spreads

What to know

Everything a microbe can do is written in its DNA. DNA is a long molecule with two strands, twisted like a ladder, and that twisted shape is the double helix. The rungs are pairs of nucleotide bases, and they pair by a strict rule. Adenine pairs with thymine. Guanine pairs with cytosine. Because the pairing never varies, one strand is a perfect pattern for rebuilding the other.

A gene is just a stretch of DNA holding the instructions for one product, usually a protein. The whole set of an organism's genetic material is its genome. Most bacteria keep their genome as one circular chromosome floating in the cytoplasm, since they have no nucleus. Many also carry small extra loops called plasmids.

Before a bacterium divides it has to make a full second copy of its DNA. That copying is replication. The helix unzips into two single strands. An enzyme called DNA polymerase reads each old strand and builds a new partner across from it, following the pairing rule. One molecule becomes two identical helices. Each new molecule keeps one old strand and one new strand, so replication is called semiconservative. Some bacteria divide every 20 minutes, and that speed is the engine behind the whole chapter. Fast copying means any gene spreads fast, and a resistance gene is no exception.

Holding the instructions is not the same as carrying them out. To use a gene the cell runs the central dogma: DNA to RNA to protein. Transcription comes first, as RNA polymerase reads a gene and writes a working copy as messenger RNA. Translation comes next, as a ribosome reads the mRNA three bases at a time. Each three-base group is a codon, and each codon names one amino acid. The ribosome links those amino acids into a protein. Change the DNA and you change the protein. Change the protein and you change what the microbe can do.

That is also why several antibiotic classes work. Fluoroquinolones block DNA gyrase, so replication stalls. Rifampin blocks RNA polymerase, so transcription stalls. Aminoglycosides, tetracyclines and macrolides jam the ribosome, so translation stalls. Each one hits a bacterial version of the machine and leaves the human version alone.

A mutation is any change in the DNA sequence. Most are copying errors. DNA polymerase is remarkably accurate but not perfect, and when a bacterium divides billions of times, rare mistakes are guaranteed. A mutation may do nothing, may harm the microbe, or once in a while may hand it a brand new ability.

The simplest change is a point mutation, where one base is swapped for another. The result varies a lot. A silent mutation changes a codon, but the new codon still names the same amino acid, so the protein comes out unchanged. A missense mutation changes one amino acid, and the protein may work a little differently or a lot differently. This is the classic route to a drug target that no longer binds its drug. A nonsense mutation turns an amino-acid codon into a stop signal, so the protein is cut short and usually breaks.

A frameshift mutation does more damage. Here a base is inserted or deleted. The ribosome reads mRNA in fixed groups of three, so adding or removing one base shifts the whole reading frame. Every codon after the change is garbled, and the protein is almost always destroyed.

Mutations that happen on their own during replication are spontaneous. The rate climbs sharply when a mutagen is present. A mutagen damages DNA or causes copying errors. Chemical mutagens include certain reactive molecules and base analogs, which slip into DNA in place of a normal base. Radiation is a powerful physical mutagen. Ultraviolet light fuses neighboring thymine bases into thymine dimers, and those kinks distort the helix. X-rays and gamma rays break strands outright.

The same damage that causes mutations will kill the cell at a high enough dose, which is why radiation is used for microbial control. Be precise about which kind. Ionizing radiation, meaning gamma rays and electron beams, goes through packaging. It is a true sterilization method for heat-sensitive supplies such as sutures, gloves and plastic syringes. UV light does not go through things. Glass, plastic, dust, organic soil and any shadow all block it, and UV does not reliably kill endospores. So UV is used for surface and air disinfection only, never sterilization.

Mutation is slow and random. If it were the only way bacteria gained new genes, resistance would spread far more slowly than it does. Bacteria have a faster second trick. They trade genes sideways. Vertical gene transfer is ordinary inheritance, from parent cell to daughters. Horizontal gene transfer, or HGT, moves genetic material between separate cells that are not parent and offspring, and they may not even be the same species. There are three routes.

Transformation is the uptake of naked DNA from the environment. When a bacterium dies and breaks apart, it spills its DNA. A nearby cell in a state called competence can pull those fragments across its membrane and stitch them into its own genome. If a fragment carried a resistance gene, the recipient is now resistant. It never had to meet the original owner alive.

Transduction uses a bacterial virus as an accidental delivery truck. That virus is a bacteriophage. When a phage infects a bacterium and builds new particles, it sometimes packs a piece of the host's DNA by mistake. That phage then infects another bacterium and injects the stray host gene along with its own.

Conjugation is the most efficient of the three, and the one that matters most for resistance. Two living bacteria make contact through a tube called a sex pilus. A copy of a plasmid then crosses from donor to recipient, and that plasmid often holds a whole packet of resistance genes. It can move genes between species that would never mix otherwise.

Besides its main chromosome, a bacterium often carries one or more plasmids. A plasmid is a small circular loop that sits apart from the chromosome, so it is extrachromosomal. The cell copies its plasmids and passes them to daughter cells, yet plasmids are not strictly required for survival. They carry bonus genes that come in handy under stress. Some let the cell use an unusual nutrient. Some build a toxin. And some let the cell survive antibiotics.

A plasmid carrying one or more resistance genes is a resistance plasmid, or R plasmid. They are dangerous for two reasons. First, one plasmid can carry resistance against several different antibiotics, so picking it up in one step makes a bacterium multidrug-resistant. Second, R plasmids often carry their own conjugation machinery, so they are built to copy themselves into neighboring cells. Put those together and you get a mobile, self-spreading package of multidrug resistance.

Not every resistant organism works this way. In MRSA the key methicillin-resistance gene sits on the chromosome, not on a plasmid. That exception is worth remembering.

How does one plasmid come to hold so many resistance genes? Often the work is done by transposons, nicknamed jumping genes. A transposon is a short DNA segment that can cut or copy itself out of one spot and paste into another. It can carry a gene from a chromosome onto a plasmid. Related elements called integrons capture resistance gene cassettes and line them up side by side. Over time these shuffling elements build a multidrug-resistance plasmid one gene at a time. Then conjugation shares the finished package.

Now the big idea. Antibiotics do not create resistance. They select for it. Picture a population of bacteria where a few cells already carry a resistance gene, either from a mutation or from a plasmid. Give an antibiotic and the susceptible cells die. The resistant few survive, divide, and refill the site. The population is now resistant from top to bottom.

Notice what did not happen. No cell changed its genes the moment the drug arrived. The drug simply cleared the field for the cells it could not kill. That is selective pressure. Every dose of antibiotic given anywhere nudges the microbial world a little further toward resistance, whether that dose was needed or not. So an antibiotic given just in case is not a harmless choice.

Resistant organisms then spread the way any microbe spreads, and that is mostly by contact: hands, equipment and surfaces. Several forces speed it up. Antibiotics get prescribed too often, including for illnesses caused by viruses. Courses get stopped early, which kills the weak cells and leaves the tough survivors behind. Heavy antibiotic use in agriculture seeds resistant organisms into the food supply. International travel carries resistant strains around the globe in days.

Slowing it down is settled where care happens, one choice at a time. Hand hygiene and contact precautions stop resistant organisms from moving between people. Antimicrobial stewardship lowers the selective pressure in the first place. Stewardship means asking whether an antibiotic is truly needed, pushing for cultures before drugs are started, narrowing broad therapy once the organism is known, and stopping antibiotics no longer indicated. Teaching matters just as much. Explain why to finish the full course. Explain why leftover antibiotics should never be shared. Explain why a cold will not respond to antibiotics.

One more consequence follows from chance mutation. If any large population will throw off a few resistant cells, then a single drug can be outlasted. That is why serious infections like tuberculosis are treated with several drugs at once.

The central dogma, drawn

The mutation types worth memorizing

Horizontal gene transfer, drawn

Sequences, in order

The central dogma runs one way, and each step has an enzyme and a drug.

1. Replication · DNA to two identical DNAs, by DNA polymerase and gyrase. Fluoroquinolones block gyrase.
2. Transcription · DNA to mRNA, by RNA polymerase. Rifampin blocks it.
3. Translation · mRNA to protein, at the 70S ribosome. Aminoglycosides, tetracyclines and macrolides block it.

Selective pressure runs as a story in four beats. Learn the beats and you can explain resistance without a diagram.

1. A population already holds a few resistant cells, by chance mutation or by plasmid.
2. An antibiotic arrives and the susceptible cells die.
3. The resistant few survive and keep dividing.
4. Their offspring refill the site, so the whole population is now resistant.

How to reason through the question types

Chapter 8Controlling Microbes: Sterilization, Disinfection and Asepsis

What to know

Here is the biggest idea in the chapter. Getting rid of germs is not one job. It is a ladder, and each rung strips away more microbial life than the rung below it.

Sterilization sits at the top. It is the complete destruction or removal of all microbial life, including the toughest target of all, the bacterial endospore. A sterile object has nothing living on it.

One rung down is disinfection. It destroys most pathogens on a non-living surface, and it does not reliably kill endospores. Because it is not absolute it comes in levels: high, intermediate and low. So much daily work is disinfection and not sterilization, and knowing the difference tells you when a wipe is enough.

The next two words get mixed up constantly. They do similar work in different places. Antisepsis is the removal of microbes from living tissue, meaning skin, a mucous membrane or a surgical site. The chemical used is an antiseptic, and it has to be mild enough to spare that tissue. Disinfection happens on non-living objects instead, and the chemical used is a disinfectant, which can be far harsher. So skin gets prepped with chlorhexidine or an alcohol swab. A counter gets wiped with bleach or a quaternary ammonium solution.

Two more words finish the set. Degerming is the mechanical scrubbing away of microbes, and the alcohol wipe before an injection does exactly that. Sanitization only lowers microbial counts to a level judged safe for public health. It is the standard for restaurant dishwashers, and it is not the standard for anything that enters the body.

Microbiology labels its weapons with two endings. An agent ending in -cidal kills the target outright. An agent ending in -static only stops the target from multiplying. The microbes are still alive and simply held in check. They will grow again if the agent is taken away, and they will grow again if the body's defenses fail. So a bacteriostatic drug leans on a working immune system to finish the job.

Now the physical methods. Heat kills in two ways at once. It denatures proteins and enzymes, and it breaks up membranes. It also comes in two forms with very different power. Moist heat means steam and boiling water, and it beats dry heat by a wide margin, because water carries energy well and makes proteins clump fast. Dry heat comes from a hot-air oven or an open flame, and it needs much higher temperatures and much longer times for the same kill.

Two old benchmarks describe what heat does. The thermal death point is the lowest temperature that kills all microbes in a suspension in ten minutes. The thermal death time is the time needed to kill them at a set temperature.

Note the key limit. Ordinary boiling at 100°C does not sterilize. It kills vegetative cells and most viruses, and it leaves hardy endospores alive. Pasteurization is a gentler cousin of boiling, using brief mild heat, and it does not sterilize either. It lowers the count of pathogens and spoilage organisms while keeping the flavor.

The autoclave is the single most important sterilizing device in health care. It beats the boiling-water problem with one trick: raising the pressure raises the boiling point, so steam can be driven well past 100°C. A standard cycle holds saturated steam at 121°C, at about 15 pounds per square inch, for roughly 15 to 20 minutes. Those conditions destroy even endospores.

The active ingredient is the moist heat of the steam. The pressure is not the thing that kills. Pressure is only there to make the hotter steam possible. A load counts as sterile only if steam reached every surface, at temperature, for long enough. So autoclaving is always verified and never assumed. Chemical indicators are the striped tape and the pouch dots that change color, and they confirm only that a load was exposed to heat. The real proof is a biological indicator: a sealed vial of Geobacillus stearothermophilus endospores. The vial rides through the cycle and is then incubated. If those famously heat-resistant spores are dead, everything else in the load is dead too.

Plenty of medical items would be wrecked by an autoclave. Plastic syringes, IV tubing, sutures and some fluids would melt, warp or break down. Those get sterilized another way. Ionizing radiation covers gamma rays, X-rays and electron beams. It carries enough energy to smash DNA and to spin off destructive free radicals. It goes deep, it works in the cold, and industry uses it for prepackaged disposables and some foods. Nonionizing radiation is mainly ultraviolet light, and it is much weaker. UV fuses neighboring thymine bases into dimers, and those dimers block DNA copying. UV barely penetrates anything, so its use is limited to disinfecting surfaces and air. It is also a real hazard to the eyes and skin of anyone standing in the room.

Filtration is different again. It removes microbes instead of killing them. A fluid or gas is strained through pores too small for cells to pass. A membrane filter with 0.2 µm pores sterilizes heat-sensitive liquids, such as certain antibiotics, vitamins and IV additives. HEPA filters do the same job for air, keeping operating rooms and protective isolation rooms clean.

Chemical agents work in a handful of ways, and most fall into two buckets. Some denature proteins, unfolding the enzymes a microbe needs. Others disrupt the plasma membrane, dissolving its lipids so the cell leaks and dies. A few damage nucleic acids or oxidize cell parts.

Whether a chemical is called a disinfectant or an antiseptic depends on where it is used, not on the molecule itself. Several agents serve as both at different strengths, and iodine and the quats are good examples.

Two practical properties decide whether a chemical is useful. The first is its use-dilution, which is the concentration at which it truly works. Stronger is not always better. 70% alcohol kills better than 100% alcohol, because alcohol needs water to denature protein. The second property is how the agent behaves around organic matter such as blood and pus, which can cancel many agents out on contact.

You do not need every compound. You do need the major families. Alcohols, meaning isopropyl and ethanol, work best at 60% to 90%. They denature proteins and dissolve membranes. They are fast, cheap skin antiseptics and surface wipes, they dry off quickly, and they do not kill spores. Chlorhexidine is a mild long-lasting skin antiseptic. It binds to skin and keeps killing for hours, which is why it is chosen for surgical scrubs and central line site care. Iodophors are iodine held inside a carrier, and povidone-iodine is the common one. They are broad-spectrum skin preps for surgery and injections.

Chlorine, which you know as household bleach or sodium hypochlorite, is strong and cheap. It is one of the few agents that reliably kills Clostridioides difficile spores on surfaces. Hydrogen peroxide and the other peroxygens oxidize cell parts and work well on surfaces and some equipment. Quaternary ammonium compounds, called quats, are low-toxicity detergent-like disinfectants for everyday surface cleaning, though they are weak against some resistant organisms.

Glutaraldehyde can sterilize delicate heat-sensitive equipment when the contact time is long enough, and so can the gas ethylene oxide. Both are toxic and both demand strict handling. Ortho-phthalaldehyde, or OPA, belongs beside them by chemistry but not by power. It is a fast-acting high-level disinfectant with poor sporicidal activity. It is cleared for high-level disinfection of endoscopes, and it is not a sterilant.

No single chemical does everything, so hospitals sort items by how much killing each one needs. That scheme is the Spaulding classification. Critical items enter sterile tissue or the bloodstream, and surgical instruments, catheters and implants all qualify. They must be sterilized. Semicritical items touch mucous membranes but not sterile tissue, and endoscopes and respiratory equipment qualify. They need at least high-level disinfection. Noncritical items touch only intact skin, and blood pressure cuffs, bedrails and stethoscopes qualify. They need only low-level or intermediate-level disinfection.

Now the arithmetic of killing. It is tempting to picture a disinfectant killing every microbe the instant it lands. Killing does not work that way. Under a lethal agent a population dies logarithmically. A constant percentage dies in each equal unit of time, not a constant number. Say a chemical kills 90% of the cells in the first minute. In the second minute it kills 90% of the survivors. In the third minute it kills 90% of those. Plot that on a log scale and you get a straight line sloping down, and that line is the microbial death curve.

The consequence runs deep. You are always killing a fraction, so you never pick off the last stubborn cell at once. You can drive the count vanishingly low, yet you can never reach true zero in a set time. That is why sterility is defined statistically, as a probability. The usual target is a sterility assurance level of one survivor in a million, written 10⁻⁶. It is not a promise that nothing is alive.

One number captures how steep that line is. It is the decimal reduction time, better known as the D-value. The D-value is the time needed to kill 90% of the population at a given temperature or agent strength. Killing 90% drops the count by one power of ten, and that drop is called one log. A short D-value means the target dies quickly. A long D-value means a tough target or a weak method. Start with a known count and a known D-value and you can work out how long a treatment must run.

The same agent can kill in seconds or fail outright, and the conditions decide which. A kill goes faster with higher temperature, with higher concentration or intensity, and with longer contact time. That last one is why every disinfectant label carries a wet contact time. The surface has to stay visibly wet for that whole time, or the agent never finishes.

A kill goes slower, or fails, under several conditions. A huge microbial load takes longer to clear, because the same percentage dies each interval. Some microbes are resistant by nature. A protective biofilm can shield them. And organic matter such as blood, pus or feces both blocks the agent physically and inactivates many agents chemically. So you always clean before you disinfect.

Microbes themselves fall into a predictable resistance ranking. Prions sit at the very top and are nearly indestructible. Bacterial endospores come next. Below them sit mycobacteria, whose waxy wall makes tuberculosis hard to kill. Then come non-enveloped viruses, then fungi. The easiest targets sit at the bottom: ordinary vegetative bacteria and fragile enveloped viruses such as influenza and HIV.

Last comes asepsis. Asepsis means the absence of contamination by pathogens, and it is practiced at two levels. Medical asepsis is also called clean technique. It aims to reduce the number and spread of microbes, and so to break the chain of transmission. It covers hand hygiene, gloves, gowns, cleaning shared equipment, and the standard precautions used with every person, whatever the diagnosis. It does not require sterility. It requires cleanliness and consistency.

Surgical asepsis is also called sterile technique. It aims for the total absence of microbes, and it is used whenever the body's natural barriers are breached: inserting a urinary or central catheter, changing a sterile dressing on a deep wound, and assisting in the operating room.

Medical asepsis rests on a small set of habits. Hand hygiene comes first. It is done before and after every contact, before a clean task, after exposure to body fluids, and after touching that person's surroundings. Alcohol-based hand rub is the default, because it is fast and effective. Soap and water are saved for hands that are visibly soiled, and they are also required for spore-formers like C. difficile. Around that core sit gloves and gowns, safe handling of linens and waste, and disinfecting shared devices between uses.

Surgical asepsis runs on a set of blunt rules, because a sterile field is only sterile until proven otherwise. Only sterile items may touch sterile items. A 1-inch border around the drape counts as contaminated. Anything held below waist level counts as contaminated. So does anything out of your line of sight, because you cannot vouch for what you cannot see. You never reach across the field, and you never turn your back on it. Any moisture that wicks through the drape contaminates it, and that wicking is called strike-through. The governing rule is honest humility. When in doubt, it is contaminated, so you discard and restart.

The death curve, drawn

The vocabulary worth memorizing

The physical methods worth memorizing

The chemical families worth memorizing

Sequences, in order

The control ladder runs from most to least. Learn it top down.

1. Sterilization · all microbial life gone, endospores included.
2. Disinfection · most pathogens on a surface, in high, intermediate or low level. Spores survive.
3. Antisepsis · the same idea, but on living tissue and with a milder chemical.
4. Degerming · mechanical removal, as with a pre-injection swab.
5. Sanitization · counts lowered to a level judged safe for public health.

The resistance ranking runs from hardest to easiest to kill.

1. Prions · nearly indestructible.
2. Bacterial endospores · the toughest living target.
3. Mycobacteria · the waxy wall makes tuberculosis hard to kill.
4. Non-enveloped viruses · norovirus and poliovirus sit here.
5. Fungi · Candida and the molds.
6. Vegetative bacteria and enveloped viruses · the easiest targets, including influenza and HIV.

Verifying an autoclave load runs in this order, and the last step is the only real proof.

1. Load so steam can reach every surface.
2. Run saturated steam at 121°C, about 15 psi, for roughly 15 to 20 minutes.
3. Read the chemical indicator tape or dot. It only shows the load met heat.
4. Incubate the biological indicator vial of Geobacillus stearothermophilus spores.
5. Dead spores mean the load is sterile. Growth means it is not.

How to reason through the question types

Chapter 9Antimicrobial Drugs

What to know

A drug that kills bacteria is a poison, and we put it into a sick person and call it medicine. One idea makes that safe. Selective toxicity means a good antimicrobial harms the microbe far more than the host. It works because a bacterium is a prokaryote, built from parts your own cells do not have. It wraps itself in a peptidoglycan wall, where your cells have none. Its ribosomes are a different size and shape. It also makes its own folic acid from scratch, while you get folate from food.

Selective toxicity is a matter of degree and never perfect. The widest safety margins belong to the drugs that attack the cell wall, meaning the penicillins and the cephalosporins. Human cells own no wall to injure. Drugs that hit something bacteria and humans share are riskier. An aminoglycoside can damage the kidney and the inner ear. An antifungal aimed at the fungal membrane can disturb human membranes too, since both are built around sterols. So the closer a microbe sits to us in biology, the harder selective toxicity gets.

Every drug has a dose that helps and a higher dose that harms. The therapeutic index puts a number on the gap. Roughly, it is the toxic dose divided by the effective dose. A large index means a comfortable margin, which is why penicillin can be given in generous doses with little worry. A small index is called a narrow therapeutic index, and now the helpful dose and the harmful dose sit uncomfortably close.

The goal is to hold the drug inside the therapeutic window: high enough to kill bacteria, low enough to spare the host. Two antimicrobials sit at the narrow end. One is vancomycin. The others are the aminoglycosides, such as gentamicin and tobramycin. With these drugs we do not guess. A trough is drawn just before the next dose, and it proves the drug has not piled up to a toxic level. Sometimes a peak is drawn as well, and it proves the level climbed high enough to work.

Note the nuance. Vancomycin also aims at the wall, yet it is still hard on the kidney and narrow-index. So the target alone does not promise safety.

Antibacterial drugs come in great variety, yet they attack a bacterium in only five places. Learn the five and almost any antibiotic drops into a mental map.

One: cell wall synthesis. The peptidoglycan wall is the pressure vessel of the cell, and without it the cell bursts. Penicillins, cephalosporins and vancomycin block the building of that wall. Growing bacteria then split open and die. Human cells have no wall, so this is the most selective target of all. The penicillin and cephalosporin families share a chemical ring called a beta-lactam, and that shared ring is why they are grouped as beta-lactam antibiotics.

Two: protein synthesis. The bacterial ribosome is a 70S machine, built from a 30S part and a 50S part. The human ribosome is 80S, built from a 40S part and a 60S part. The sizes differ, so a drug can jam the bacterial machine and leave yours running. Aminoglycosides and tetracyclines hit the 30S part. Macrolides such as azithromycin and erythromycin hit the 50S part, and so does clindamycin. The match is not perfect, though. Your mitochondria carry ribosomes that look bacterial, and that overlap helps explain some of the toxicity these drugs cause.

Three: nucleic acid synthesis. Some drugs block the enzymes bacteria use to copy and unwind DNA. The fluoroquinolones work this way, and ciprofloxacin and levofloxacin are the common ones. They shut down bacterial DNA gyrase and topoisomerase, which are shaped differently from the human versions.

Four: metabolic folate pathways. Bacteria have to build their own folic acid, while you eat yours. Sulfonamides and trimethoprim starve the bacterium of folate by blocking two steps of that pathway, one after the other. It is a target you simply do not possess.

Five: cell membrane. A few drugs punch holes in the bacterial membrane, and the contents of the cell then leak out. These drugs tend to be harsher, because your cells have membranes too, so they are held back for stubborn infections. Polymyxins such as colistin attack the outer membrane of Gram-negative rods. Daptomycin works the other way around. It wrecks the membrane of Gram-positives such as MRSA and VRE, and it cannot touch a Gram-negative at all.

Drugs also differ in how thoroughly they act. A bactericidal drug kills the bacterium outright, and penicillins, aminoglycosides and fluoroquinolones do that. A bacteriostatic drug only halts growth, and tetracyclines, macrolides and sulfonamides work that way. They hold the population still and leave the finishing blow to the immune system.

In a healthy person, static is often enough. It is not always enough. Think of someone whose defenses are down, such as a person on chemotherapy with a low neutrophil count. Think of endocarditis or meningitis, where few immune cells reach the site. In those cases the team reaches for a bactericidal drug, because there is no reliable immune backup to mop up the survivors.

A drug's spectrum of activity is the range of bacteria it works against. A narrow-spectrum drug hits only a few kinds. A broad-spectrum drug hits many at once. Broad sounds better, and in a crisis it can save a life. Breadth still has a cost. A broad-spectrum antibiotic clear-cuts the normal flora along with the pathogen, and that opens the door to a superinfection. The two common ones are Clostridioides difficile colitis and a yeast overgrowth. Every needless exposure also nudges the whole population toward resistance. So the aim, once the culprit is known, is the narrowest drug that will do the job.

Cultures take time, and a person with sepsis cannot wait a day or two. So therapy usually happens in two stages. The first stage is empiric therapy, which is an educated guess. The guess rests on three things: where the infection probably started, what the local resistance patterns look like, and how sick the person is right now. The team starts a broad drug at once, but only after cultures are drawn. Then the lab reports what grew and what it is susceptible to, and the team switches to definitive therapy. Ideally that means narrowing to a focused drug, a move called de-escalation. So the order at the bedside matters. Cultures first. Antibiotics second.

How does the lab know which drug to recommend? It tests the isolate against a panel of antibiotics. The key number is the minimum inhibitory concentration, or MIC, the lowest drug concentration that stops the bacterium from growing. Several methods get at it. The classic one is the Kirby-Bauer test, also called disk diffusion. Paper disks soaked in different antibiotics sit on a lawn of bacteria, and a clear ring called the zone of inhibition forms where growth was stopped. The bigger the zone, the more susceptible the organism.

Note what disk diffusion does and does not give you. It sorts the isolate into susceptible, intermediate or resistant. It does not read out an exact MIC. The Etest strip sharpens this into a real MIC value, and automated broth microdilution trays do the same job in tiny wells. Either way the result reaches the chart as a letter: S, I or R.

Resistance is not the drug wearing off. It is evolution in fast forward. The drug kills the susceptible bacteria and leaves the rare survivor that happened to carry a defense. There are four main defenses.

Destroy the drug. The famous example is beta-lactamase, an enzyme that snips open the beta-lactam ring before the drug can act. Penicillinase in Staphylococcus aureus was the first one found. Gram-negative rods have taken the idea further. ESBLs chew through the cephalosporins, and carbapenemases such as KPC chew through the carbapenems. This is also why some penicillins come paired with a beta-lactamase inhibitor, and amoxicillin-clavulanate is one such pair. The inhibitor soaks up the enzyme so the antibiotic survives.

Change the target. If the drug can no longer recognize its binding site, it cannot act. Change the lock and the key stops working. This is the trick behind MRSA. The bacterium picks up a gene for an altered penicillin-binding protein that beta-lactams cannot grip. VRE does the same thing to vancomycin, swapping the end of its wall building block so the drug has nothing to clamp onto.

Pump the drug out. Efflux pumps sit in the membrane and bail the antibiotic back out before it reaches a lethal level inside the cell. Some pumps move several unrelated drug classes, so one pump can cause broad resistance. Tetracycline resistance often works this way.

Keep the drug out. A Gram-negative cell gates its outer membrane with channels called porins. Narrow those channels, or lose them, and less drug gets in at all. Paired with a pump, the level inside never gets high enough to kill.

Resistance can start with a plain mutation. Its frightening speed comes from horizontal gene transfer. Bacteria hand resistance genes to one another, even across species, and the genes travel on mobile loops of DNA called plasmids. One plasmid can carry resistance to several drug classes at once, so a single exposure can select for a strain resistant to drugs it never met.

MRSA is methicillin-resistant Staphylococcus aureus. It resists essentially every beta-lactam antibiotic, which knocks out the safest and most familiar drugs on the shelf. The team is pushed toward vancomycin and its relatives instead. MRSA spreads readily in hospitals and in the community, so people known to carry it are often placed on contact precautions. It does not travel alone. VRE means vancomycin-resistant enterococci. ESBL-producing and carbapenem-resistant Gram-negatives sit on the same list, and so does multidrug-resistant tuberculosis.

Antibiotic stewardship is the shared effort to use these drugs only when they help, and to use them well. It is often boiled down to the five rights: the right drug, the right dose, the right route, the right duration, and the right person. The purpose runs two ways. One aim is a better outcome for the person in front of you, with fewer side effects, less C. difficile and a shorter stay. The other aim is to protect the drugs themselves for the people who come after.

At the bedside stewardship means drawing cultures before the first dose. It means giving antibiotics on time, so blood levels stay effective. It means watching for improvement and for toxicity. And it means asking the questions that trim needless use. Can this IV antibiotic switch to a pill, now that the person is eating and getting better? Is this still needed on day five, or is it just still ordered? Are two drugs covering the same organism? One more job belongs here: the last check on the allergy history. Clearing up a vague penicillin allergy keeps someone off a broader, harsher drug for no reason.

Stewardship reaches past the hospital door, and out there teaching is the tool. Explain that antibiotics do nothing for colds, flu and most sore throats. Teach three habits. Take the drug exactly as prescribed. Never save or share leftover antibiotics. Never use someone else's. The old blanket rule to always finish every pill is being refined, because some infections do better on a shorter course. So the honest, durable message is simpler: take it the way it was prescribed, and call before stopping early or doubling up.

The drug classes worth memorizing

The resistance mechanisms worth memorizing

Sequences, in order

Choosing a drug runs in a fixed order, and the first step is the one most often skipped.

1. Draw cultures before the first dose.
2. Start empiric therapy, usually broad, based on likely source, local resistance and how sick the person is.
3. Read the lab report: the organism, the count and the S, I or R susceptibilities.
4. Switch to definitive therapy, narrowing to a focused drug. That move is de-escalation.
5. Review the duration, switch IV to oral when the person is stable and eating, and stop when it is no longer needed.

Resistance builds in a sequence too, and each step is faster than the one before.

1. A chance mutation or an inherited gene gives one cell a defense.
2. Selective pressure from a drug kills the susceptible cells and leaves that one to multiply.
3. Horizontal gene transfer, above all conjugation, hands the gene to other cells and other species.
4. An R plasmid gathers several resistance genes, so one transfer confers multidrug resistance.
5. Contact spread on hands, equipment and surfaces moves the organism from person to person.

How to reason through the question types

One pageCram sheet